Written by the people who actually build and support these systems. No vendor spin, and no article that ends by telling you the answer is always more software.
Right-sizing, reserved capacity, egress charges and the scheduled job that quietly costs more than the servers. What to check before you migrate, and what to check ninety days after.
Read moreFour automation patterns with a genuine payback period — and three popular ones that mostly generate impressive demos and disappointing quarterly reviews.
Read moreRole-based access, audit logging, secret management, dependency scanning and three more — all cheap on day one, all expensive to retrofit in year three.
Read moreFlutter and React Native have closed most of the gap — but not all of it. The specific cases where native still wins, and how to decide without relitigating it every sprint.
Read moreThe strangler pattern explained without jargon, using a fifteen-year-old VB6 inventory system that is still running today — one module at a time.
Read moreStaff do not resist new systems out of stubbornness. They resist systems that make their day harder. What we changed in our process after one rollout went badly.
Read moreVersioning, pagination, error shapes, idempotency and rate limits — the five decisions that determine whether your API is a pleasure or a support burden.
Read moreHow to cut an MIS screen down to the six metrics management will actually act on — and where to put everything else so nobody feels it was lost.
Read moreA practical scoping method for founders: how to reach something credible in ten weeks without building an architecture you will have to throw away at scale.
Read moreNew articles are published roughly twice a month. Subscribe in the footer to get them by email.
Articles only go so far. A 30-minute call with an engineer will get you further than a week of research.